Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science




Goda, Katsuichiro

2nd Supervisor


Joint Supervisor


The Mw 7.8 2012 Haida Gwaii Earthquake triggered a tsunami that highlighted the importance of tsunami hazard assessment on Canada’s Pacific coast. Stochastic source modelling serves as a valuable method to assess future tsunami hazard and has not been performed for this region. The source models characterize the uncertainty of earthquake ruptures by considering variability in fault geometry and slip heterogeneity, which, in turn, allows the consideration of a wide range of tsunami scenarios in the Haida Gwaii region. The model predictions are constrained by observational data and past source inversion studies. One hundred twenty-eight stochastic tsunami scenarios are generated using the stochastic source modelling method to assess tsunami hazard via tsunami inundation simulations of the target region and conduct sensitivity analyses of tsunami height variability. The resulting models can promote better-informed risk management decisions and future probabilistic tsunami hazard analysis in this region.

Summary for Lay Audience

On October 28, 2012, an Mw 7.8 earthquake hit the region of Haida Gwaii, Canada. The tsunami triggered by the earthquake was recorded across the Pacific Ocean. Horizontal and vertical deformations were obtained months after the earthquake and, during post event field surveys, run-up levels were measured at several locations within the rupture zone. This study conducts a tsunami analysis of the Haida Gwaii region using stochastic source modelling and performs Monte Carlo tsunami simulation to develop source models that generate tsunami waves in close match with the recorded observations. The developed stochastic earthquake source model can be applied to evaluate tsunami hazards due to future tsunamigenic events in Haida Gwaii. The methodology encompasses the wavenumber analysis of six existing earthquake slip models to define a generic fault model for the synthetic slip source generation. The stochastic source parameters are based on earthquake source scaling relations derived from global models. The stochastic method uses spectral synthesis, where key slip characteristics are specified in slip statistics, slip distribution parameters, and asperity areas. For a given set of stochastic synthesis parameters, slip distributions are generated by a Fourier integral method. The derived stochastic models can capture realistic asperity zones and source parameters close to those of the 2012 event. Asperity zones are mainly located on the shallow ocean side of the fault, which is consistent with the epicentre location constrained by seismic and deformation data. Consequently, simulated tsunami waves at different stations show that first wave amplitudes are in agreement with the observations. Simulated tsunami run-ups are generally consistent with those observed at sites sheltered from storm waves, with differences ranging from 0.5−3 m. In contrast, the differences become significant at sites exposed to storm waves with a discrepancy of up to 7 m. The discrepancy may be attributed to the possibility that run-up survey observations at exposed bays might include effects due to major storm events that hit Haida Gwaii between the earthquake and the survey. Moreover, source parameters and models that are calibrated for the 2012 event can be adopted to evaluate tsunamis due to future large events in the region.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.